(IJCSIS) International Journal of Computer Science and Information Security, Vol. 7, No. 2, February 2010 INDEXER BASED DYNAMIC WEB SERVICES DISCOVERY Aihab Khan, Malik Sikandar Hayat Khiyal Saba Bashir, M.Younus Javed,Farhan Hassan Khan Fatima Jinnah Women University National University of Science & Technology Rawalpindi, Pakistan Rawalpindi, Pakistan , Abstract-Recent advancement in web services plays an With the increasing technology it is becoming more critical important role in business to business and business to to find a service that satisfies user requirements. Service consumer interaction. Discovery mechanism is not only discovery is a mechanism used to find the appropriate service used to find a suitable service but also provides that meets desired criteria. Three main roles in web services collaboration between service providers and consumers by are providers, consumers and registries. Service providers using standard protocols. A static web service discovery publish their web services in registries and clients discover mechanism is not only time consuming but requires their desired services by querying them using SOAP message continuous human interaction. This paper proposed an protocol [1]. efficient dynamic web services discovery mechanism that In manual discovery process a human uses a discovery can locate relevant and updated web services from service service, which involves the input of user requirements and registries and repositories with timestamp based on through matchmaking process, output the desired results [2]. indexing value and categorization for faster and efficient The main obstacle in this mechanism is heterogeneity problem discovery of service. The proposed prototype focuses on [3]. Different types of heterogeneities are: quality of service issues and introduces concept of local • Technological heterogeneities (different platforms or cache, categorization of services, indexing mechanism, different data formats). CSP (Constraint Satisfaction Problem) solver, aging and • Ontological heterogeneity (difference in domain specific usage of translator. Performance of proposed framework terms and concepts). is evaluated by implementing the algorithm and • Pragmatic heterogeneity (different development of correctness of our method is shown. The results of domain-specific processes) proposed framework shows greater performance and This makes the technology limited in terms of efficiency accuracy in dynamic discovery mechanism of web services and availability. Whereas in dynamic discovery process an resolving the existing issues of flexibility, scalability, based agent performs the above task either at design or run time, but on quality of service, and discovers updated and most there are different constraints of this process for example relevant services with ease of usage. interface requirements, standardization of protocols and trust of stakeholders etc. This paper focuses on the dynamic web Keywords-component: Classification of services; Dynamic services discovery approach. discovery; Indexing; UDDI registry; Web services. With the increasing number of web services, it is becoming more difficult to manage description on distributed environment such as on web. Key limitations of current I. INTRODUCTION approaches for dynamic web services discovery are: 1. Lack of fully dynamic web services discovery As an enabling technology, web services are software mechanism. components that are used to present services on internet. Each 2. Lack of mechanism that solves heterogeneity problems. web service has standard interface through which it 3. Clients cannot learn from past experiences to improve communicates with other web services in order to provide future decisions. different functionalities. Service providers register their web 4. Need of mechanism which store and retrieve up to date services in registries, typically in UDDI (Universal information on consumer’s request. Description Discovery and Integration). HTTP (Hyper Text 5. Need of mechanism which gives in time response of Transfer Protocol) protocol is used for communication of query in case of searching services from web. SOAP (Simple Object Access Protocol) messages exchanged between service providers and consumers. 153 http://sites.google.com/site/ijcsis/ ISSN 1947-5500 (IJCSIS) International Journal of Computer Science and Information Security, Vol. 7, No. 2, February 2010 6. In existing approaches, clients have to search from web UDDI is used in discovery process where keywords are (registries) each time when they request for services matched by using intelligent algorithms. The research is XML which is time consuming task and requires lot of effort. schema matching where various string comparison techniques 7. Service providers and consumers don’t use same language are applied like suffix, prefix, and infix for keyword matching. for registration and discovery. This technique is useful where many acronym, jargons etc are Keyword based discovery is one possible way to discover used in web services [8]. the web services. The advantage of UDDI is that it filters and Liang-Jie Zhang, Qun Zhou [9] proposed a dynamic web ranks services quickly. The drawback is that it contains only services discovery framework for web services representation metadata. It is based on keyword matching and it has only chain. It solves the manual search process problems in linked single search criteria. In some cases it may not produce the documents. The multiple linked documents are searched using desired results due to the words that are semantically related WSIL (Web Services Inspection Language) chains and results with each other [2]. are aggregated and returned them to users all at once. It also In order to take a full advantage of web services, user must re-explores the nested WSIL documents by using pre-fetched discover and invoke them dynamically. Semantic web services link calculation and caching methods. are a possible solution for automated web service discovery in Sudhir Agarwal [10] proposed a goal specification language which services are semantically described and accessed. It to discover the web services. It describes that the constraints resolves the heterogeneity as well as interoperability problems. should be specified on functional and non functional Ontologies are used for manipulation of data. There are two properties in goal specification language. Then this goal drawbacks of semantic web services discovery approach. First, specification language is used to discover the desired services. it is impossible for all service providers to publish their The language is novel combination of SHIQ(D) and µ- services in same ontology. Second, already existing web calculus. It covers resources, behavior and policies of agents services does not have associated semantics and it is not involve in discovery process. Java API (Application possible to convert all of them into same ontology [4]. Programming Interface) is used to describe goals and concrete syntax is used so that end users can easily write the formulas A. Contribution through interface. This research paper proposed a framework that presents Fu Zhi Zhang et al. [2] presented an algorithm for dynamic solutions to most of the problems in dynamic web service web services discovery. The proposed algorithm is based on discovery. More precisely, it uses the translator to convert composition of web services with the process model. It is service requests and responses in one form that can be easily based on OWL-S (Ontology Web Language-Semantic) used by the system. Indexer is used to index services to ontology. Unlike previous proposed models for dynamic simplify search process. CSP solver is used to select most discovery it can have multiple executions and matchmaking appropriate service. Aging factor helps to maintain updated for a single service request and return desired service. information. The discovery process is fastened by the use of Paul Palathingal [11] presented an agent based approach for local cache which allows clients to discover services based on service discovery and utilization. Agents acts instead of users previous experiences. and dynamically discover, invoke and execute web services. The organization of paper is such that Section 2 describes Through this technology sender object must not know the the previous research on dynamic services discovery. In receiver’s address. Also the interoperability between section 3, some basic information about service discovery and distributed services is being achieved through these software dynamic service discovery is given. Section 4 presents the components. The selected agents match the desired query from detailed overview of proposed technique including framework their respected repositories and send the retrieved results to and key mechanisms. Section 5 describes the working and composition agent who composes the web services and sends algorithms of proposed framework. Section 6 presents the back to requester. implementation and analysis of proposed framework. Finally, Stephan Pöhlsen and Christian Werner [12] proposed a the conclusion and future work is given in section 7. dynamic web services discovery framework in large networks. It solves the discovery problem of mobile ad-hoc networks by II. RELATED WORK introducing a Discovery Proxy (DP). Through Discovery proxy clients can easily search desired web services through Web services are XML (Extensible Markup Language) service providers and get updated information as services does based software